DEBT POLICY AS CORPORATE GOVERNANCE MECHANISM IN CONCENTRATED OWNERSHIP

This research aims to test the debt policy as a mechanism to reduce agency conflict among majority and minority shareholders. This test aims to answer the problems to what extend debt can be used as corporate governance mechanism in a sense of reducing agency conflict. This research is important since most of company ownership structure in Indonesia is categorized concentrated structure, where its make a conflict between majority and minority shareholders. The populations of the research are companies that go public in the Indonesian capital market until the year of 2003. These samples of this research consist of 40 companies that are selected based on nonprobability technique with purposive sampling method. They were then divided into two groups, high concentrated ownership structure and low concentrated ownership structure. In processes testing the hypothesis, 2 indicators were used, i.e. market indicator and accounting indicator. Event study analysis was used for market indicator, whereas multiple regression analysis was used for accounting indicator. Based on empirical examination result, it is generally concluded that debt policy cannot be used as mechanism to reduce agency conflict among majority and minority shareholders, both at high and low concentrated ownership structure. This is because of average company debt are higher than average industry debt. Debt policy tend be used as a tool of expropriation to minority shareholders. The expropriation is higher at high concentrated ownership structure rather than at low concentrated ownership structure and the difference is significant.


Research Background
The main issue of this research is the fact that the ownership structure in Indonesia is categorized high concentrated, where the current agency conflict is the conflict between majority shareholders with minority shareholders. Agency conflict occurs because of asymmetric information and different objective among stakeholders in the company. Agency conflict occurs between shareholders and managers or between shareholders and creditor (Jensen and Meckling, 1976). But later, agency conflict also occurs between majority shareholders and minority shareholders, between majority shareholders and other stakeholders, like employees and suppliers (Zhuang et al., 2000;and Ariyoto, 2000). This pohcy is effective as corporate governance mechanism, when market response this policy positively. It will lead to increase company performance (Denis, 2001).
Debt policy is used as corporate governance mechanism to reduce agency conflict (Jensen and Meckling, 1976;McConnel and Servaes, 1995;Faccio et al, 2001). The increasing of debt can reduce conflict of tree cash flow and show to public that majority shareholders do not use the free cash flow for their own sake. The increasing of debt will drive a company to use the cash efficiently, because the cash is used to pay debt interest periodically. Debt generates external monitoring; consequently, the majority shareholders should conduct the best performance to improve the company's performance. This is referred as control hypothesis (Faccio et al., 2001;Jensen, 1986;Sarkar and Sarkar, 2005).
The structure of ownership determines agency conflict type. At dispersed ownership structure, the agency conflict that might occur is the conflict between shareholders and managers. Nevertheless, at concentrated ownership structure, the agency conflict that might occur is the conflict between majority shareholders with minority shareholders. The ownership structure in Indonesia is high concentrated. Therefore, the researcher finds it interesting to analyze the influence of corporate governance on the company's performance by using debt policy analysis on the concentrated ownership structure. The hypothesis of this research is whether the debt policies can be used as corporate governance mechanism that can influence the company's performance. The objective of this research is to examine the influence of corporate governance mechanism on the company's performance by using debt policy analysis. The motivations of this research are as follow. First, The fact that the concentrated ownership structure in Indonesia leads to the agency conflict between majority shareholders with minority shareholders drives the researcher to examine debt policy as corporate governance mechanism on the levels of high and low concentrated ownership structures. Throughout this analysis, the researcher can understand the appropriate policies for each concentrated ownership structural level. Second, the research results concerning the influence of policy of debt to company's performance still contradict to one another.
This research is expected to give the contribution to empirical contribution, methodologies, and pohcies. Empirical contribution is to prove the debt policy as corporate governance mechanism to reduce agency conflict among majority and minority shareholders. The result of the research is expected to support the reference books on corporate governance mechanism in correlation with the concentrated ownership structure in emerging market and to be the resource for the coming research related to corporate governance. Methodologies contribution indicates that so far there have been no researches examining the policy to reduce agency conflict in the concentrated ownership structure using market indicator and accounting indicator. The existing previous research examined only one policy using one indicator. Therefore, the result of this research is expected to give information to BAPEPAM and stock exchange concerning the implementation of corporate governance in Indonesia capital market, especially from policy of debt of view. By doing so, it can be of advantage as a reference in making or completing corporate governance regulation in correlation with minor shareholders rights and ownership structure in Indonesian capital market. The result of this research is to give information for companies in Indonesian capital market related to the implementation of good corporate governance.

Literature Review and Hypothesis Development 2.1. Corporate Governance and Debt
The concept of corporate governance is derived from agency theory. Agency theory explains the appearance of conflict, the essence of conflict, and also solution to the conflict. Agency theory (Jensen and of Meckling, 1976) states that there is a separation between ownership and control in modern companies. This separation generates the agency conflict because of its asymmetric information from agency and principal. Agency conflict also appears in the existence of free cash flow in a company, referred as free cashflow hypothesis (Jensen, 1986). Nevertheless, since the problem of agency progressively becomes complex, corporate governance is needed. The definition of corporate governance in general is a system, structure, mechanism or policy, process and also rules explaining the relations between all parts in a company, so that they are able to carry out the rights and obligations correctly and proportionally. There are 2 paradigms of corporate governance; shareholding paradigm and stake-holding paradigm (Letza and of Sun, 2002). There are 4 principles of corporate governance (Gregory and of Simms, 2000), i.e., fairness, transparency, accountability, and responsibility. The effectiveness of corporate governance is determined by some factors: ownership structure, law and enforcement, economy system, social, culture, process, and also clear performance measurement.
Debt policy is used as corporate governance mechanism to reduce agency conflict (Jensen and Meckling, 1976;McConnel and Servaes, 1995;Faccio et al., 2001). The increasing of debt can reduce conflict of free cash flow and show to public that majority shareholders do not use the free cash flow for their own sake. The increasing of debt will drive a company to use the cash efficiently, because the cash is used to pay debt interest periodically. Debt generates external monitoring; consequently, the majority shareholders should conduct the best performance. This is referred as control hypothesis (Faccio et al., 2001;Jensen, 1986;Sarkar and Sarkar, 2005). Ownership structure determines agency conflict type. At dispersed ownership structure, the agency conflict that might occur is the conflict between shareholders and managers. Nevertheless, at concentrated ownership structure, the agency conflict that might occur is a conflict between majority shareholders with minority shareholders.

Market Indicator Hypothesis Formulation
Concentrated ownership structure causes agency conflicts between majority and minority shareholders (Shleifer and Vishny, 1997;Zhuang et al., 2000). This agency conflict occurs because of the asymmetric information from majority shareholders and minority shareholders. Besides, majority shareholders have bigger power to control the managers in decision making, for example, the one related to the company's free cash flow. Debt can be used to decrease agency conflict between majority and minority shareholders, because debt allows public to notice that majority shareholders do not use the free cash flow for themselves, but it is used to pay the debt and interest periodically. Debt shifts management monitoring from shareholders to creditors (Jensen and Meckling, 1976;Jensen, 1986;Faccio et al., 2001). This monitoring forces the management or shareholders to conduct actions, which can give, benefit to the company. This is called control hypothesis. Nevertheless, excessive debt will decrease the company's performance, because the increase of debt will be followed by the increase of debt expense. -

Hi:
Market reacts positively to the bond announcement at concentrated ownership structure Ownership structure determines agency conflict type. At concentrated ownership structure, the agency conflict that might happen is the conflict between majority shareholders with minority shareholders (Shleifer and Vishny, 1997;Zhuang, et al., 2000). Majority shareholders have power to control the manager so that decision made on free cash flow is intended for their own benefit rather than for minority shareholders. This complies to Shleifer and Vishny (1997)'s statement saying that when concentrated ownership comes to a certain limit, the majority shareholders can fully control the company and they tend to make policies that give benefit to themselves. The level of concentrate structure influences the level of agency conflict between majority and minority shareholders. At high concentrated ownership structure, agency conflict is higher than that at low concentrated ownership structure (Gugler and Yurtoglu, 2000;Dewenter and Warther, 1998). This means, the higher concentrated ownership structure, the higher the power owned by majority shareholders to expropriate minority shareholders.

H2:
Market react positively greater toward the bond announcement at : ' high concentrated ownership structure to low concentrated ' -r ownership structure

Accounting Indicator Hypothesis Formulation
Ownership structure in a company determines power distribution between all parts in a company. Therefore, at concentrated ownership structure, majority shareholders have very big power to influence managers in making decision, thus the decision made will give benefit only for them but will gain loss for minor shareholders. Therefore, the agency conflict that might occur is the conflict between majority shareholders and minority shareholders (Shleifer and Vishny, 1997;Zhuang et al., 2000). The ownership structure determines agency conflict type. At high concentrated ownership structure, the agency conflict that might happen is the conflict between majority shareholders with minority shareholders. This conflict occurs due to the fact that the structure of ownership is mainly composed of family or founders' members who have big power in controlling managers in decision making. Therefore, the decision made tends to give benefit only for themselves on minor shareholders' account. This statement is proven by Mitton (2002) stated that when majority shareholders is entangled in management as a director or commisaris, they will have an opportunity or bigger incentive to expropriate minority shareholders.
Debt policy can be used as a corporate governance mechanism to decrease agency conflict in a company (Jensen and Meckling, 1976;McConnel and Servaes, 1995), because debt can decrease the tree cash flow in a company. Debt also allows public to notice that majority shareholders do not use the free cash flow for themselves; rather it is used to pay the debt and interest periodically. Debt shifts management monitoring fi-om shareholders to creditors (Jensen and Meckling, 1976;Jensen, 1986;Faccio et al., 2001). This monitoring force the management or shareholders conduct actions, which do not disadvantage minority shareholders, and consequently, it gives positive influence to the company's profitability. This is called control hypothesis (Faccio et al, 2001;Jensen, 1986;Sarkar and Sarkar, 2005).

V
H3: Debt to total asset ratio has positive influences toward the profitability of concentrated ownership structure The level of concentrated ownership structure determines the agency conflict in a company. The higher the concentrated ownership structure, the bigger the agency conflict between majority shareholders and minority shareholders. This comphes with Shleifer and Vishny (1997)'s statement saying that when concentrated ownership comes to a certain limit, the major shareholders can control the company and they tend to make policies that give benefit to themselves. The higher the concentrated ownership structure, the bigger the majority shareholder's power to expropriate minority shareholders. Therefore, debt policy at high concentrated ownership structure has bigger positive influence than that at low concentrated ownership structure on the company's profitability.
H4: Debt to total asset ratio on high concentrated ownership structure influences greater positively than low concentrated ownership structure at the company's profitability

Research Methods
The population is all companies listed in Indonesian stock market until 2003. The sample was gained through non-probability technique with purposive sampling method; with the following criteria: a) nonfinancial companies listed in Indonesian stock market; b) companies whose shares are owned by the largest shareholders, with a minimum ownership as high as 20%. This complies with BAPEPAM law that defines majority shareholders as those who owns at least 20% shares; c) the company announces of obligation. A method of data pooling is also used to give a better analysis result. Based on the above criteria, 40 samples achieved are then divided into two groups: i.e. a) low concentrated ownership structure (low COS), where a company's shares are owned by the largest shareholders from 20% to less than 50%; b) high concentrated ownership structure (high COS), where a company's shares are owned by the largest shareholders as many as 50% or more.

Market Indicator Testing.
Market performance is measured by abnormal return dan cumulative abnormal return (Husnan, 2001;Gugler and Yurtoglu, 2000 Abnormal return is an excess between actual return with expected return. Average abnormal return is observable when event is announced (t = 0). Cumulative average abnormal return being tested is the one with t = -2 until t = +2 and t = 0 until t = + 5. Abnormal return is measured by using single index market model with an estimation period of 50 days (t = -60) and (t = -11). For market indicator, debt is indicated by obligation right issue announcement.

Accounting Performance Testing
Accounting performance is measured by return on equity. Diagnostic examine is carried out to test the multikolinieritas and heteroscedasticity. Multikolinieritas test is carried out to test whether the independent variables have one or more linier relation. To test the multikolinieritas problem, tolerance value or variance inflation factors test is conducted. Heteroscedasticity test is carried out to detect whether (cr2) variance from dependent variable is increasing as a result of the increase in independent variable. To detect the heteroscedasticity test, Glejser testing is conducted (Gujarati, 2003). Based on the testing with accounting indicators, independent variables are debt. Debt is indicated by leverage = total debt/total assets. Company size is used as control variable. ROE = ai + PiOwn + p2 Debt + p4 (Own x Debt) + where: ROA = Return On Assets Own = 1 for high concentrated ownership Own = 0 for low concentrated ownership Debt = total debt/total asset , '

*Significant at a = 5%
The data used is secondary data: consist of, annual report from the year 1996 -2003, obligation announcement, daily stock price, daily stock price index, and other information related with this research. The data are compiled from ICMD, JSX Statistics, PRPM, PDBI, Bisnis Indonesia daily, internet, stock market data from PPA-UGM, and other sources related to the research. To what extent how far the debt policies can be use as corporate governance mechanism, the researcher is should test the significance of the values of AAR and CAAR on 4 groups, i.e. a) the cash flow increases as the investment opportunity set is high; b) the cash flow increases as the investment opportunity set is low; c) the cash flow decreases as the investment opportunity set is high; d) the cash flow decreases as the investment opportunity set is low.
The 4 groupings above explain that in discussing the hypothesis of free cash flow, the starting point is not on how to measure free cash flow, rather on how to make a decision on cash flow when faced with investment opportunity set. The agency problem of free cash flow occurs when the increasing cash flow is faced with low investment opportunity set; this is hypothesis of free cash flow (Jensen, 1986). The low investment opportunity clearly shows that a company does not have a positive net present value. This condition indicated that management to use excessive cash for inefficient things that will give disadvantages to shareholders. The bigger the free cash flow in a company is, the bigger its flexibility owned by a company is. This situation can also lead the management to flexibly use the existing cash; where the chances of any interest conflict in using the free cash flow is higher as well. It is the reason as to why it is important to lessen the free cash flow; for instance, by increasing the payment of dividend, debt, and investment. The result will support the research hypothesis, if the median of average abnormal return and cumulative average return on the high cash flow with low investment opportunity is positive or negative and is statistically significant.

Result and Discussion 4.1. Result on Market Indicator
Hypothesis 1 states that market react positively to bond announcement. Table 1 indicated that the values of AAR, CAAR2 and CAAR5 are negative and statistically significant. This condition indicates that the increase in debt cannot be used as corporate governance mechanism in a company. Total debt exceed will decrease share price. This result does not support Jensen and Meckling (1976) and Jensen (1986). Hypothesis 1 is accordingly rejected. That negative reaction must be proved later. Table 2 indicated at row 2 columns 3, CAAR2 and CAAR5 negative and statistically significant. So, cash flow increase and investment opportunity set low, agency conflict high and bond announcement react negatively by market, this is support to rent extraction hypothesis.
The negative response is possibly due to the relatively big amount of debt exceeding the maximum value. The amount of debt below the maximum value can function as a monitoring tool and can also be used to increase a company's performance. However, once the amount exceeds the maximum value, the debt will lead to high agency cost and bankruptcy. In the end, this will diminish a company's performance. Hypothesis 2 states that market react greater positively on high concentrated ownership structure to low concentrated ownership structure. (2.1830)* (-0.8050) (-0.5830) 2. Low Concentrated -0.0120 -0.0069 -0.0123 (-7.6317)* (-1.4900) (-1.7790) * Significant at a = 5%; AAR = Average Abnormal Retum indicated market react to obligation announcement at 2; CAAR2 = Cumulative Average Abnormal Retum indicated market react to obligation announcement at announcement until 2 days after announcement; CAAR5 = Cumulative Average Abnormal Retum indicated market react to obligation announcement at announcement until 5 days after announcement. Table 3 show that the values of AAR, CAAR2 and CAAR5 at high concentrated ownership are negative and statistically significant. The same result applies to low concentrated ownership that shows the negative and statistically significant value of AAR. This condition indicates that debt policy cannot be used as corporate governance mechanism on high and high concentrated ownership.

Table 2. Relationship Cash Flow and Investment Opportunity
That negative reaction at high concentrated ownership must be proved later. Table 4 indicated at row 2 columns 3, CAAR2 and CAAR5 negative and statistically significant. Therefore, cash flow increase and investment opportunity set low, agency conflict high and bond announcement react negatively by market, this is support to rent extraction hypothesis.
Negative reaction at low concentrated ownership must be proved later. Table 5 indicated at row 2 columns 3, AAR and CAAR5 negative and statistically significant. Therefore, cash flow increase and investment opportunity set low, agency conflict high and bond announcement react negatively by market, this is support to rent extraction hypothesis. In general, public companies in Indonesia add more debt without adding their own asset as guarantee. This condition will burden the debt holders and shareholders since this may lead to bankruptcy. Based on the calculation, it turns out that the comparative ratio between the debt amount and the asset amount increased from 1999 to 2003; i.e. = 0.49216; 0.51162; 0.57432; 0.51730; and 0.57568. This statement corresponds with Taridi (1999) who states that companies in Indonesia have high debt. The capability to expropriate increases at the same time as the debt excessively increases (Harris and Raviv, 1988).  The result is consistent with Faccio et al. (2003); Taridi (1999), Haris and Raviv (1988);Faccio et al. (2001); also with Sarkar and Sarkar (2005) who state that any debt on concentrated ownership structure will bring about moral hazardous attitudes that endanger negatively on a company's performance. On this kind of concentrated ownership structure, shareholders have the power to expropriate minor shareholders; referred as expropriation hypothesis. Faccio et al. (2003) state that in developing countries with the characteristics of concentrated ownership like Indonesian, debt cannot fimction as a monitoring tool to lessen any agency conflict; rather it will serve as a tool of expropriating minor shareholders by major shareholders

Table 4. Relationship Cash Flow and Investment Opportunity Debt at High Concentrated Ownership
The reasons as to why expropriating through debt on a concentrated ownership structure is possible are: a) the protection on the minor shareholders is weak. These are proven by Alba, Claessens, and Djankov (Taridi 1999) who state that Indonesia is among countries in East Asia whose protection on the minor shareholders is weak; b) Indonesian stock market has not yet so well developed that debt cannot yet function as an effective corporate governance mechanism; c) the fact that a company's reputation is still dominated by majority shareholders indicates that the company still has its intrinsic weakness. This is understandable since once the headquarter files a bankruptcy due to excessive debt, there will be difficulties as to who should be responsible simply because the control system is complicated in a pyramidal structure (Faccio et al.,2001).

Result on Accounting Indicators
The result of diagnostic examine shows that regression model has multicolinearity problem. Multicolinearity is not a serious problem if the aim of analysis is to predict. It is simple due to the fact that the higher the R2, the better the prediction will be (Gujarati, 2003). The testing result shows that model regression suffer some heteroscedasticity. This problem can be overcome by weighted least square.
Hypothesis 3 states that ratio of debt to total asset has positive influence on a company's profitability. The testing result of model regression shows that the value of debt coefficient is negative (-77.586) and statistically significant. This means that debt cannot be used as corporate governance mechanism in a company. This result does not support hypothesis 1.
Hypothesis 4 states that ratio of debt to total asset on high concentrated ownership structure has bigger impact compare to low concentrated structure does on the company's profitability. The regression value of model shows that the value of debt coefficient on high concentrated ownership structure is negatively smaller than that on low concentrated ownership structure. (-77.586 < -77.586 + 56.961) and statistically significant. It means that the influence of debt on high concentrated ownership structure is negatively smaller than that on low concentrated ownership structure on the company's performance. This is hypothesis is not supported.
Debt is proven to have negative impact on a company's performance either in low or high concentrated ownership structure. This is due to: a) in general, Indonesian companies are funded by debts; moreover, any raise in debt is without any increase in asset as a guarantee. This existing condition will only make debt holders as well as shareholders pay the price of any possible bankruptcy. Based on the calculation, a comparative ratio between the amount of debt and asset is relatively increasing starting from the year 1999 until 2003; i.e. 0.49216; 0.51162; 0.57432; 0.51730; and 0.57568. This statement corresponds to Taridi (1999) who states that companies listed in Indonesian stock market have big amount of debt. The capability to expropriate is increasing as excessive debt increases at the same time (Harris and Raviv, 1988). In other words, any debt in certain amount will function as a monitoring tool so as to help increase a company's performance. However, once the amount of debt is way beyond a maximum level, the debt will only diminish a company's performance; b). concentrated ownership structure impels major shareholders to expropriate minor shareholders. This is likely to occur since its structure is so pyramidal that minor shareholders will find it difficult to control any conducts of major shareholders. Third, Indonesian Stock Market has not yet well developed and enforcement towards corporate governance rules is relatively low. This will also indicated major shareholders to expropriate the minor shareholders.

Conclusion
Base on the result of general empirical test can be concluded that debt poHcy cannot effective be used as reducing agency conflict between majority and minority shareholder, both on high and low concentrated ownership structure. This is because average debt on high and low concentrated ownership structure are greater than average debt industry. Greater expropriation is existed on high ownership concentrated than low ownership concentrated.
This research has limitations as follows. First, determining t = 0 correctly is a difficult task so that any incorrect t = 0 determination will only weaken the research result. Second, concentrated ownership proxy being used in the research is the maximum ownership by an individual or institution with minimum ownership in different concentrated ownership; for instance by using herfmdahl index, institution, family, and so on. By using different proxies, it is hoped to clearly reveal the role of ownership structure in corporate governance. Third, it is not sufficient to use only secondary data in finding out the effectiveness of corporate governance mechanism in increasing a company's performance. Thus, the following research should be better to use the primary data in order to get better result.
This research has some implications. For academic purpose, this research is beneficial as foundation of conducting further researches, especially for those who want to develop corporate governance. It is worth to note that the success of corporate governance mechanism depends more on some factors and the existence of relations among variables. BAPEPAM and the capital market need to evaluate their regulations and to increase the quality of enforcement related to corporate governance since there are still a lot to do in implementing corporate governance. So far, any practices on corporate governance are just merely acts of practicing the law without any concern on how to make it a necessity for the sake of the company and the investors.